Author: brian
Description:
We are using XHTML 1.0 Transitional - I presume we are using valid XHTML. If so, we
should have an option in the software to add an XHTML icon, which individual wikis
can customize using skins.
I suggest having an option to not display the icon because if the icon is always
there, it will be downloaded even if not displayed.
Version: unspecified
Severity: enhancement
URL: http://uitest.com/en/analysis/
robchur wrote:
If people want one, they will add it. We're already adding a "powered by
MediaWiki" icon, and we don't really want to clutter up the UI.
avarab wrote:
The parser is not guarenteed to output valid XHTML, see open parser bugs and
some parser tests which fail.
avarab wrote:
One could make an extension that did this however, and If one wanted to do it
properly one would pass the output through some validator (for instance the w3c
html validator which is freely avalible) before displaying it, REOPENING and
putting it into extension requests.
brian wrote:
(In reply to comment #1)
If people want one, they will add it. We're already adding a "powered
byMediaWiki" icon, and we don't really want to clutter up the UI.
Why don't we want one? I thought XHTML was part of an effort to reduce the
computational power required to view pages, which makes information more
accessible, which is the point of this foundation.
brian wrote:
The new URL has links to 2 XHTML validators, including the W3C validator, at the
time of writing and is linked to from both [[en:HTML]] and [[en:XHTML]].
rowan.collins wrote:
(In reply to comment #5)
Why don't we want one? I thought XHTML was part of an effort to reduce the
computational power required to view pages, which makes information more
accessible, which is the point of this foundation.
But do we need a button saying "this is XHTML" in order to gain this
accessibility? And is determining which pages are and which aren't valid
actually useful, other than for finding weak points in the code? Or would it be
better to work on guaranteeing that the output was *always* valid, whether or
not we put a human-readable blob in the UI saying so.
brian wrote:
No, we don't need that icon. I am suggesting that it would promote XHTML,
however. We promote the organisation and the software (with icons), so why not
promote the language - isn't it just as important?
Yes, we should work on guaranteeing that the output is *always* valid. However,
is this a realistic aim? We have not succeeded yet.
rowan.collins wrote:
(In reply to comment #8)
No, we don't need that icon. I am suggesting that it would promote XHTML,
however. We promote the organisation and the software (with icons), so why not
promote the language - isn't it just as important?
I'm not sure it is as important, really. After all, the canonical version of our
content is in an ad-hoc non-standardised syntax defined only by the program it
has "evolved" in tandem with.
Yes, we should work on guaranteeing that the output is *always* valid. However,
is this a realistic aim? We have not succeeded yet.
Until we do so, though, isn't almost hypocritical to "promote" XHTML - i.e.
encouraging others to obtain a standard we have failed to reach.
I mean, I'm not dead against doing any of this; it just seems rather a waste of
time/effort.
robchur wrote:
(In reply to comment #5)
Why don't we want one? I thought XHTML was part of an effort to reduce the
computational power required to view pages, which makes information more
accessible, which is the point of this foundation.
...well the design of the icon might not fit with the user's skin, for starters.
Slapping images about for the sake of it is a Bad Idea. To my mind, it's a
rather cosmetic change. As Rowan says, it seems like a waste of time and effort.
brian wrote:
(In reply to comment #9)
...so why not promote the language - isn't it just as important?
I'm not sure it is as important, really. [snip]
Really? Without the language, isn't the content no more useful than a Chinese
text to someone who only knows English?
Until we do so, though, isn't almost hypocritical to "promote" XHTML - i.e.
encouraging others to obtain a standard we have failed to reach.
The question is not whether we have reached this standard but rather how hard we
are trying to do so.
brian wrote:
(In reply to comment #10)
...well the design of the icon might not fit with the user's skin, for starters.
That could be said of any icon.
Slapping images about for the sake of it is a Bad Idea.
It's not for the sake of it.
To my mind, it's a rather cosmetic change.
Won't some people say that for any icon?
As Rowan says, it seems like a waste of time and effort.
Having the "valid XHTML" icon will help promote the language, as explained above.
It will also help find bugs in the software. How else are people finding the
XHTML issues?
rowan.collins wrote:
(In reply to comment #11)
Really? Without the language, isn't the content no more useful than a Chinese
text to someone who only knows English?
But our content isn't *in* XHTML, it's in an ad hoc markup language we call
"wikitext"; without a proper parser (which arguably we have not yet got), it is
indeed like being in a foreign language. Our imperfect "translation" into XHTML
may be an *attempt* to make it easier to process, but it's not really anything
to boast about right now.
Anyone wanting to machine-process our content right now is as likely to play
with the wikitext as the XHTML output; once we've got tools that can
consistently produce meaningful XML (and other) formats, it will be worth
pointing these out to people.
As a debugging tool, I see a little more merit, but it may be that a properly
structured wiki2xml parser (as started by Magnus Manske) will make all those
issues redundant anyway.
brian wrote:
(In reply to comment #13)
But our content isn't *in* XHTML
As far as most readers are conerned, it is. Most readers never see the wikitext.
avarab wrote:
Why is this extension request RESOLVED LATER? There's no reason why it couldn't
be implemented now...
brian wrote:
(In reply to comment #15)
Why is this extension request RESOLVED LATER? There's no reason why it couldn't
be implemented now...
If I had to guess, I would say that "resolution" indicates we intend to fix it
eventually, but we have more important things to do first.
But then no one would ever see it, as it was resolved, so how would we remember?
rowan.collins wrote:
(In reply to comment #14)
But our content isn't *in* XHTML
As far as most readers are conerned, it is. Most readers never see the wikitext.
True. I guess what I was thinking was that most of our *reusers* (to whom we're
promoting our reusability) get the wikitext and do something with that (though
this may well involve running it through MediaWiki's "parser" to produce some
form of HTML). Ideally, they'd be able to get something more standard, and as I
say the most likely avenue for this is Magnus's XML parser, for which there will
hopefully be front-ends for things like DocBook and ODF as well as XHTML.
I still think it seems weird to talk of "promoting" a standard we're not sure we
actually adhere to.
brian wrote:
(In reply to comment #17)
I still think it seems weird to talk of "promoting" a standard we're not sure we
actually adhere to.
But I am sure we are making significant progress. Some people promote things that
don't actually exist yet.